Molecular Neurobiology https://doi.org/10.1007/s12035-018-0996-x

Filamentous Aggregation of Sequestosome-1/p62 in Brain Neurons and Neuroepithelial Cells upon Tyr-Cre-Mediated Deletion of the Atg7

Supawadee Sukseree1 & Lajos László2 & Florian Gruber1,3 & Sophie Bergmann1 & Marie Sophie Narzt1,3 & Ionela Mariana Nagelreiter1,3 & Romana Höftberger4 & Kinga Molnár2 & Günther Rauter5 & Thomas Birngruber6 & Lionel Larue7,8,9 & Gabor G. Kovacs4 & Erwin Tschachler1 & Leopold Eckhart 1

Received: 28 July 2017 /Accepted: 7 March 2018 # The Author(s) 2018

Abstract Defects in autophagy and the resulting deposition of aggregates have been implicated in aging and neurodegenerative diseases. While gene targeting in the mouse has facilitated the characterization of these processes in different types of neurons, potential roles of autophagy and accumulation of protein substrates in neuroepithelial cells have remained elusive. Here we report that Atg7f/f Tyr-Cre mice, in which autophagy-related 7 (Atg7) is conditionally deleted under the control of the tyrosinase promoter, are a model for accumulations of the autophagy adapter and substrate sequestosome-1/p62 in both neuronal and neuroepithelial cells. In the brain of Atg7f/f Tyr-Cre but not of fully autophagy competent control mice, p62 aggregates were present in sporadic neurons in the cortex and other brain regions as well in epithelial cells of the choroid plexus and the ependyma. Western blot analysis confirmed a dramatic increase of p62 abundance and formation of high-molecular weight species of p62 in the brain of Atg7f/f Tyr-Cre mice relative to Atg7f/f controls. Immuno-electron microscopy showed that p62 formed filamentous aggregates in neurons and ependymal cells. p62 aggregates were also highly abundant in the ciliary body in the eye. Atg7f/f Tyr-Cre mice reached an age of more than 2 years although neurological defects manifesting in abnormal hindlimb clasping reflexes were evident in old mice. These results show that p62 filaments form in response to impaired autophagy in vivo and suggest that Atg7f/f Tyr-Cre mice are a model useful to study the long-term effects of autophagy deficiency on the homeostasis of different neuroectoderm-derived cells.

Keywords Autophagy . Protein aggregation . Sequestosome-1 . p62 . Cortex . Ependyma . Choroid plexus

Electronic supplementary material The online version of this article (https://doi.org/10.1007/s12035-018-0996-x) contains supplementary material, which is available to authorized users.

* Leopold Eckhart 5 Division of Biomedical Research, Medical University of Graz, [email protected] Graz, Austria

6 1 Research Division of Biology and Pathobiology of the Skin, Joanneum Research, Health - Institute for Biomedicine and Health Department of Dermatology, Medical University of Vienna, Sciences, Graz, Austria Lazarettgasse 14, 1090 Vienna, Austria 7 Institut Curie, INSERM U1021, CNRS UMR3347, Normal and 2 Department of Anatomy, Cell and Developmental Biology, Eötvös Pathological Development of Melanocytes, PSL Research Loránd University, Budapest, Hungary University, Orsay, France 3 Christian Doppler Laboratory on Biotechnology of Skin Aging, 8 INSERM,Orsay,France Vienna, Austria 4 Institute of Neurology, Medical University of Vienna, 9 Equipe labellisée – Ligue Nationale contre le Cancer, Université Vienna, Austria Paris 11, Orsay, France Mol Neurobiol

Introduction cells with Tyr promoter activity leads to the permanent inacti- vation of Atg7. When these cells proliferate, Atg7-dependent Autophagy is a mechanism for the delivery of cell components autophagy remains suppressed in all progeny cells. Atg7f/f Tyr- to lysosomes for hydrolytic degradation. The main type of au- Cre mice show mild hypopigmentation of hair and tail skin [18] tophagy is macroautophagy which involves the formation of but otherwise appear phenotypically normal. Autophagy is also double-membraned vesicles, known as autophagosomes, suppressed in the retinal pigment epithelium of Atg7f/f Tyr-Cre around substrates. A set of autophagy-related , such as mice leading to the accumulation of p62 and an increase in the Atg5 and Atg7, is essential for this process, and deletion of these abundance of a degradation-prone variant of retinal pigment genes suppresses autophagy in mice [1, 2]. Adaptor epithelium-specific 65 kDa protein (RPE65) [20]. such as sequestosome 1, also known as p62 [4], differentially The characterization of mice carrying the Tyr-Cre trans- bind to autophagy substrates and introduce specificity into the gene has shown that Cre expression and Cre-mediated gene degradation process [3]. Autophagy removes many types of deletions do not only occur in pigment cells but also in distinct protein aggregates, dysfunctional organelles, and other poten- groups of neurons of the developing brain [21, 22]. tially dangerous cell components but also contributes to the Specifically, Tyr-Cre-mediated gene deletions were reported recycling of macromolecules to ensure cellular homeostasis in the basal forebrain, hippocampus (dentate gyrus pyramidal [3, 5–8]. cell layers), olfactory bulb, the granule cell layer of the lateral Sequestosome-1/p62 is a multifunctional protein compris- cerebellum cortex, sympathetic cephalic ganglia, ing domains that bind to the mammalian Atg8 homolog leptomeninges of the telencephalon, and cranial nerves (V), microtubule-associated protein 1 light chain 3 (LC3), which (VII), and (IX) [22]. By contrast, the neuroepithelial cells of mediates docking of autophagy substrates to the forming the adult brain such as the ependyma and the choroid plexus autophagosome, ubiquitinated proteins, and the Nrf2 regulator epithelium have not been reported to be affected by Tyr-Cre- Keap1 [4]. Via its N-terminal Phox and Bem1p (PB1) domain, mediated DNA recombination [21, 22]. p62 is able to self-oligomerize in the form of filaments [4]. Here we investigated Atg7f/f Tyr-Cre mice for p62 accumu- Suppression of autophagy results in the intracellular accumu- lations signifying suppression of autophagy in non-pigment lation of p62 [4]. p62 is present in neurofibrillary tangles in cells. We show that p62 accumulates in neuroepithelial cells of Alzheimer’s disease and Lewy bodies in Parkinson’sdisease the ocular ciliary body, the choroid plexus and the ependyma [9]. Together with reports about the decline of autophagic as well as in neurons of the brain. By immunogold labeling activity in aged organs and impaired clearance of and electron microscopy, the ultrastructure of these p62 ag- autophagosomes in neurodegenerative diseases, aberrant pro- gregates is revealed to consist of filaments both in neurons and cessing of p62 in diseased tissues has suggested a particularly neuroepithelial cells. Our data establish Atg7f/f Tyr-Cre mice important role of autophagy in the aging brain [10–12]. as a model for the study of aging-associated aberrant p62 Accordingly, the pharmacological inducer of autophagy, depositions in cells of the neuroectodermal lineage. rapamycin, has been suggested as therapeutic agent for aging-associated neurodegeneration [13, 14]. Cell types of different functions and turnover rates vary in Material and Methods their dependence on autophagy for the elimination of damaged organelles and potentially harmful protein aggregates as well as Mice for recycling of building blocks of macromolecules [3, 8, 15]. While constitutive deletion of either Atg5 or Atg7 leads to peri- The generation and maintenance of Atg7f/f Tyr-Cre mice have natal lethality in mice [16, 17], cell type-specific deletions of been reported previously [18]. Briefly, Atg7f/f mice (kindly autophagy genes via the Cre-loxP system allows to inactivate provided by Masaaki Komatsu, Tokyo Metropolitan Institute autophagy in a targeted manner and to determine whether lack of Medical Science, Tokyo, Japan) were crossed with mice of autophagy plays essential roles in these specific cells [1, 2]. carrying the Tyr-Cre transgene [21]. Tissue samples were pre- In previous studies, we have generated Atg7f/f Tyr-Cre mice pared from age-matched Atg7f/f Tyr-Cre and Atg7f/f mice. Only for the investigation of the role of autophagy in pigment cells hemizygous males and homozygous females were included to [18–20]. The Tyr-Cre gene utilizes promoter and enhancer ele- avoid possible effects of X inactivation on the ments from tyrosinase (Tyr), a gene encoding the enzyme that Tyr-Cre transgene in heterozygous females [23]. converts tyrosine to melanin via tyrosine hydroxylase and dopa oxidase catalytic activities in pigment cells. The Tyr promoter Immunohistochemical and Immunofluorescence drives the expression of a transgene encoding the Cre Analysis recombinase, which deletes the region between two loxP sites. The target sites have been introduced into an essential part of For histological investigations, the eyes were enucleated im- the autophagy gene Atg7 [17], so that the expression of Cre in mediately after sacrificing mice. Likewise, the brain and other Mol Neurobiol tissues were prepared. The tissue samples were fixed in were counterstained with uranyl acetate and lead citrate prior 4% paraformaldehyde over night and then embedded in to investigation with a JEOL JEM-1011 electron microscope. paraffin. Thin-sections were investigated by immunohis- tochemistry and immunofluorescence labelling according to published protocols [24] with modifications. The sec- Western Blot Analysis tions were incubated with polyclonal rabbit anti-Sqstm1/ p62 (MBL International Corporation, dilution, 1:1000) Brains were lysed in a protein extraction buffer containing followed by incubation with goat anti-rabbit immunoglob- 50 mM Tris (pH 7.4), 2% SDS, and complete protease inhibitor ulin conjugated to horseraddish peroxidase for 30 min. In cocktail (Roche, Mannheim, Germany) and homogenized by immunofluorescence double labelings, anti-p62 was used sonication. The insoluble debris was removed by centrifuga- besides mouse monoclonal anti-tyrosine hydroxylase tion, and the protein concentration of the supernatant was mea- (Millipore, MAB318, clone LNC1, 1: 400) and mouse sured by the bicinchoninic acid (BCA) method (Pierce, monoclonal anti-ubiquitin (Millipore, 1:500). The follow- Rockford, IL). Western blot analysis was performed as de- ing secondary antibodies were used for immunofluores- scribed previously [20]. Twenty microgram protein was loaded cence labeling: goat anti-rabbit immunoglobulin coupled per lane on SDS polyacrylamide gels (ExcelGel SDS, gradient to Alexa-Fluor 488 (green) or Alexa-Fluor 546 (red) 8–18, Amersham Biosciences) on a horizontal electrophoresis (Molecular Probes, Leiden, The Netherlands), and goat system (Amersham Biosciences) and thereafter blotted onto a anti-mouse immunoglobulin coupled to Alexa Fluor 546 nitrocellulose membrane. For the detection of specific antigens, (Life Technology, 1:500). Counterstaining of nuclei was the following first step antibodies were used: rabbit polyclonal done with hematoxylin for immunohistochemistry and anti-p62 (BML-PW9860-0100, Enzo Life Sciences, NY, dilu- Hoechst 33258 (Molecular Probes) for immunofluores- tion 1:2000), rabbit anti-Atg7 (Sigma, 1:1000), and mouse anti- cence analysis. Isotype antibodies of unrelated specific- GAPDH (HyTest Ltd., Finland, 1:2000). As secondary anti- ities were used instead of the primary antibodies in nega- bodies, goat anti-rabbit immunoglobulin G (IgG) (Bio-Rad tive control experiments. The labeled sections were Laboratories, CA) and sheep anti-mouse immunoglobulin G photographed under a fluorescence microscope using the (GE Healthcare, UK) antibodies conjugated to horseradish per- Metamorph software. oxidase were used at a dilution of 1:10000. The bands were revealed with enhanced chemiluminescence reagent (ThermoFisher Scientific). Immunogold Labeling and Electron Microscopy

Whole brains were immersely fixed in immune fixative contain- RNA Preparation, Reverse Transcription, ing 3.2% paraformaldehyde, 0.2% glutaraldehyde, 1% sucrose, and Quantitative PCRs and 3 mM CaCl2 in 0.1 M Na-cacodylate buffer for overnight incubation at 4 °C. Pieces of 3 × 3 × 3 mm of the lateral ventricle RNA was prepared from brains using the RNeasy Plus wall were resected. The small tissue blocks were cryoprotected Mini kit (Qiagen, Hilden, Germany) according to the in 30% sucrose in Na-cacodylate for 24 h. The blocks were manufacturer’s instructions. RNA was reverse-transcribed frozen in liquid nitrogen and subsequently transferred to anhy- using the iScript cDNA synthesis kit (Bio-Rad drous methanol containing 0.5% uranyl-acetate at − 70 °C. After Laboratories, Hercules, CA) according to the manufac- 6 h, the temperature was raised to − 20 °C and the dehydration turer’s protocol. Quantitative real-time PCRs with was continued for 24 h with gentle agitation. Then specimens SYBR-Green in the LightCycler system (Roche Applied were infiltrated with pure LR Gold at − 20 °C for 24 h (three Science, Mannheim, Germany) were performed according incubationsof8heach)andthenpolymerizedfor96hat− to a published protocol [18, 19]. Transcripts of the fol- 20 °C using a DL-103 12 W ultraviolet lamp. lowing genes were amplified with the indicated primers: Ultrathin sections were collected on formvar film-coated Beta-2 microglobulin (B2m) (Mm_B2m_f, 5′- nickel grids. For epitope retrieval and quenching, the samples attcacccccactgagactg-3′ and Mm_B2m_r, 5′ - were treated with 0.3% Na-borohydride in Tris-buffered saline tgctatttctttctgcgtgc-3′), γ-glutamyl cystine ligase modula- containing 50 mM NH4Cl and 50 mM glycine for 10 min at tory subunit ( Gclm)(Mm_Gclm_f,5′ - room temperature. After antigen retrieval, the samples were tggagcagctgtatcagtgg-3′ and Mm_Gclm_r, 5′- incubated with affinity-purified rabbit polyclonal anti-p62 an- agagcagttctttcgggtca-3′), NAD(P)H:quinone oxidoreduc- tibody (1:100 dilution, overnight at 4 °C), followed by incu- tase 1 (Nqo1) (Mm_Nqo1_f, 5′-gaagctgcagacctggtgat-3′ bation with goat anti-rabbit immunoglobuin secondary anti- and Mm_Nqo1_r, 5′-ttctggaaaggaccgttgtc-3′), and Sqstm- body conjugated with 10 nm gold particles (Sigma Aldrich, 1/p62 (Mm_p62_f, 5′-ccagtgatgaggagctgaca-3′ and 1:100, 6 h at room temperature). The immuno-labeled sections Mm_p62_r, 5′-tgggcacacactgcacttat-3′)[18]. Mol Neurobiol

Preparation and Quantification of Lipids Student’s t test. P values below 0.05 were considered significant. Mouse brain tissue (n = 4 per genotype) was homogenized in the ninefold volume of methanol/acetic acid (3%)/butylated Ethics Statement hydroxytoluene (BHT, as antioxidant, 0.01%). Samples were purified using the liquid-liquid extraction procedure [25]and Mice were maintained and sacrificed by cervical dislocation were reconstituted in 85% aqueous methanol containing according to the animal welfare guidelines of the Medical 5 mM ammonium formate and 0.1% formic acid. Analysis University of Vienna, Austria, as approved by the Ethics was performed at FTC-Forensic Toxicological Laboratory, Review Committee for Animal Experimentation of the Vienna. Aliquots (10 μl) were injected onto a core-shell type Medical University of Vienna, Austria, and the Federal C18column(Kinetex2.6μm, 50 mm × 3.0 mm ID; Ministry of Science, Research and Economy, Austria (GZ Phenomenex, Torrance, CA) kept at 20 °C and using a 1200 66.009/0255-II/3b/2013). CSF was prepared under approval series HPLC system (Agilent, Waldbronn, Germany), coupled of the Federal Ministry of Science, Research and Economy, to a 4000 QTrap triple quadrupole linear ion trap hybrid mass Austria (BMWFW-66.010/0045-WF/V/3b/2015). All spectrometer system with a Turbo V electrospray ion source methods were performed in accordance with the relevant (Applied Biosystems, Foster City, CA, USA). Elution was guidelines and regulations. performed according to a published protocol [25]. Detection was carried out in positive ion mode by selected reaction monitoring (SRM) of 99 MS/MS transitions using product Results ion (m/z 184), the diagnostic fragment for the phosphocholine residue. Data acquisition and instrument control were per- Tyr-Cre-Mediated Deletion of Atg7 Leads formed with Analyst software, version 1.6 (Applied to Accumulation of p62 in Neurons Biosystems). Individual values were normalized to the intrin- and Neuroepithelial Cells of the Brain sic 1,2-di-palmitoyl-3-phosphorylcholine (DPPC) for brain extracts. Non-oxidized native lipid species (1-palmitoyl-2- As the Tyr-Cre transgene has been reported to be active not arachidonoyl-sn-glycero-3-phosphorylcholine (PAPC) m/z only in pigment cells but also in other cells of the 782; 1-palmitoyl-2-linoleoyl-sn-glycero-3-phosphorylcholine neuroectodermal lineage [21, 22], we analyzed the brain of (PLPC) m/z 758; 1-stearoyl-2-arachidonoyl-sn-glycero-3- Atg7f/f Tyr-Cre mice and, as control, the brain of fully phosphorylcholine (SAPC) m/z 810; 1-stearoyl-2-linoleoyl- autophagy-competent Atg7f/f mice. Both male and female sn-glycero-3-phosphorylcholine (SLPC) m/z 786) and chain Atg7f/f Tyr-Cre mice were successfully used in breeding, had fragmented oxidized species (1-palmitoyl-2-(5-oxovaleroyl)- normal weight, and could be kept up to an age of 2 years. sn-glycero-3-phosphorylcholine (POVPC) m/z 594; 1- However, Atg7f/f Tyr-Cre mice older than 1.5 years showed palmitoyl-2-azelaoyl-sn-glycero-3-phosphorylcholine abnormal limb-clasping reflexes that were also reported in (PAzPC) m/z 666; 1-stearoyl-2-azelaoyl-sn-glycero-3- mouse models of neurodegenerative diseases [1, 2, 28] phosphorylcholine (SAzPC) m/z 694 1-palmitoyl-2- (Suppl. Fig. S1). To determine possible changes in the central glutaroyl-sn-glycero-3-phosphorylcholine (PGPC) m/z 610; nervous system of Atg7f/f Tyr-Cre mice, brains of young (age 1-palmitoyl-2-(oxo-nonanoyl)-sn-glycero-3- 1–2 months) and old (age 14–27 months) mice of both geno- phosphorylcholine (PONPC) m/z 650 and 1-stearoyl-2-(oxo- types (Atg7f/f Tyr-Cre and Atg7f/f mice) were investigated. nonanoyl)-sn-glycero-3-phosphorylcholine (SONPC) m/z Immunohistochemistry showed that p62 was present only 678) were identified as isobaric and co-eluting with commer- at minimal amounts in the brain of Atg7f/f mice whereas it cial standards [25]. accumulated in various areas of the brain of Atg7f/f Tyr-Cre mice (Fig. 1). The abundance of p62 was massively increased Preparation of CSF and Analysis of Proteins in 2-year-old Atg7f/f Tyr-Cre mice (Fig. 1) but accumulations by Electrophoresis of p62 were already present in 1-month-old mice of this ge- notype (Suppl. Fig. S2). Neurons in the cortex (mainly in the The CSF was prepared according to a published protocol [26]. frontal and parietal lobe, large spherical inclusions in the py- Proteins were separated by polyacrylamide gel electrophoresis ramidal layer), the basal ganglia, parts of the thalamus, and and subjected to silver staining [27]. few neurons in the brain stem contained p62 in the form of aggregates with diameters of up to 12 μm(Fig.1). Besides Statistics p62 accumulations in cell bodies of these regions, round to oval p62 deposits were also present in neuronal processes. By The statistical significance of differences between sample contrast, the hippocampus, leptomeninges, dentate nucleus, groups was examined using the two-tailed unpaired and the cerebellum of Atg7f/f Tyr-Cre mice were Mol Neurobiol

Fig. 1 Tyr-Cre-mediated deletion Atg7f/f Atg7f/f Tyr-Cre of Atg7 leads to differential B accumulation of p62 in cells of the A brain. p62 was immunohistochemically (IHC) stained (red) in the brain of Atg7f/f IHC: p62 (A,C)andAtg7f/f Tyr-Cre (B, D) mice aged at least 14 months. In negative control experiments the anti-p62 antibody was replaced by equally concentrated antiserum of unrelated specificity (E, F). The images show the C D cortex (A, C) and the brain regions around the third ventricle ependyma (C–F). Scale bars, 50 μm ependyma IHC: p62 choroid plexus choroid plexus

E F

IHC: neg. control

immunonegative for p62. In the substantia nigra, tyrosine response such as Nqo1 and Gclm1 were not upregulated in the hydroxylase-positive cells which produce neuropigment did brain of Atg7f/f Tyr-Cre mice (Suppl. Fig. S4). not contain accumulations of p62 (Suppl. Fig. S3). Although there are, to the best of our knowledge, no reports about Tyr- Atg7f/f Tyr-Cre-Induced Aggregates of p62 Are Not Cre-mediated recombination in neuroepithelial cells of the Strictly Associated with Ubiquitin brain [21, 22], we detected accumulations of p62 in the ependyma and the choroid plexus epithelium of Atg7f/f Tyr- Immunofluorescence analysis confirmed that p62 aggregates Cre mice (Fig. 1). were present both in cell bodies and dendrites of many neu- Western blot analysis confirmed the strong increase in p62 rons in Atg7f/f Tyr-Cre mice (Fig. 3). Besides aggregates of abundance in the brains of Atg7f/f Tyr-Cre mice relative to that diameters in the range of 0.5–2 μm, large aggregates of up to of Atg7f/f mice (Fig. 2A). While the level of p62 was below the 10 μm in diameter were detected in 2-year-old mice (Fig. 3K). Western blot detection limit in brains of Atg7f/f mice, p62 was A fraction of ependymal and choroid plexus epithelial cells consistently detected at the expected size and in the form of contained elevated amounts of p62 and p62 bodies in Atg7f/f high molecular weight protein species, indicating oligomeri- Tyr-Cre but not in Atg7f/f mice (Fig. 3N, Q). The p62 aggre- zation, in the brain of Atg7f/f Tyr-Cre mice. Western blot anal- gates in neuroepithelial cells reached sizes similar to those in ysis did not show a difference of total Atg7 amounts in the brain neurons (up to 8 μm in the ependyma and up to 12 μmin brain lysates of the two mouse genotypes (Fig. 2B), which the choroid plexus). Interestingly, most of the p62 aggregates was consistent with the finding that the vast majority of cells were present in the apical cytoplasm of epithelial cells of the were not altered with regard to Atg7-dependent p62 degrada- choroid plexus (Fig. 3Q) whereas p62 bodies appeared to be tion in the brain of Atg7f/f Tyr-Cre mice (Fig. 1). In contrast to randomly distributed in ependymal cells of Atg7f/f Tyr-Cre the elevation of p62 at the protein level, Sqstm1/p62 mRNA mice (Fig. 3N). Notably, only a subset of epithelial cells in was not increased in the brain of Atg7f/f Tyr-Cre mice (Suppl. the ependyma and choroid plexus contained p62 aggregates Fig. S4), suggesting that the accumulation of p62 protein was whereas the others were immunonegative for p62. not driven by enhanced biosynthesis but by reduced degrada- Immuno-labeling for ubiquitin showed that some but not tion. Targets of the autophagy-sensitive Nrf2-mediated stress all p62 bodies in the brains of Atg7f/f Tyr-Cre mice were Mol Neurobiol

A showed that p62 was concentrated in aggregates in the cyto- plasm of neurons (Fig. 4A, C) and ependymal cell at the ven- f/f Tyr-Cre Tyr-Cre Tyr-Cre tricle wall (Fig. 4B, D) of Atg7 Tyr-Cre mice whereas no or f/f f/f f/f f/f f/f f/f f/f only sparse p62 labels were detected in the brain of Atg7 Mw mice (up to an age of 26 months). The aggregates in neurons Atg7 Atg7 Atg7 Atg7 Atg7 Atg7 [kD] and ependymal of Atg7f/f Tyr-Cre mice were composed of electron-dark filaments that were densely decorated with HMW anti-p62 immunogold labels (Fig. 4C–F). The aggregates p62 170 were not surrounded by a membrane. The organization of 130 the p62 bodies was similar in all affected brain cell types 100 investigated and in mice aged 9 months (Suppl. Figs. S6, p62 70 S7) and 23 months (Fig. 4). 55 Tyr-Cre-Mediated Deletion of Atg7 Is Associated 40 with an Increase in Dicarboxylic Acid-Containing 35 Phospholipids in the Brain 25 Autophagy contributes to the lipid metabolism of cells [30, 15 31], and we have recently identified oxidized phospholipid B species that accumulated in autophagy-deficient melanocytes Atg7 70 that acquired a premature senescent phenotype [19]. We thus C applied a recently developed HPLC-MS/MS method to inves- tigate the abundance of selected oxidized phospholipid spe- 40 GAPDH cies in brains from Atg7f/f and Atg7f/f Tyr-Cre mice. The azelaic acid-containing oxidation products of two major un- Fig. 2 Western blot analysis shows increase of p62 abundance and saturated phospholipds, PLPC and SLPC, 1-palmitoyl-2- formation of high-molecular weight species of p62 in the brain of Atg7f/f Tyr-Cre mice. Protein lysates from whole brains of Atg7f/f and azelaoyl-sn-glycero-3-phosphocholine (PAzPC), and 1- Atg7f/f Tyr-Cre mice (n = 3 per genotype, age: 23–26 months) were stearoyl-2-azelaoyl-sn-glycero-3-phosphocholine (SAzPC) analyzed by Western blot for p62 (A), Atg7 (B), and GAPDH (C). were significantly increased relative to the saturated internal Positions of molecular weight markers are indicated on the right. control lipid, 1-,2-dipalmitoyl-3-phosphocholine (DPPC) HMW, high molecular weight; kD, kilo-Dalton; Mw, molecular weight (Fig. 5), which was similar to the increase of PAzPC in cul- tured autophagy-deficient melanocytes of Atg7f/f Tyr-Cre mice associated with immunoreactivity for ubiquitin (Fig. 3). Only [19]. By contrast, the levels of the unoxidized PLPC and in few neurons, mainly located in the cortex (Fig. 3J–L), the SLPC as well as those of Lyso-PPC and Lyso-SPC were not relative increase of ubiquitin was as pronounced as that ob- significantly different between Atg7f/f and Atg7f/f Tyr-Cre mice served for p62, whereas many cells with p62 accumulations (Fig. 5). Of note, phospholipid-esterified azelaic acid can bind did not have increased levels of ubiquitin. to lysine residues [32] and thereby may contribute to Electrophoretic analysis suggested that the cerebrospinal lipoxidative damage and aggregation of proteins in cells af- fluid (CSF), which is secreted by the choroid plexus epitheli- fected by Tyr-Cre-induced suppression of autophagy. um and, to a smaller extent, by the ependyma [29], contained the same major protein species at the same relative abundance f/f f/f p62 Accumulates in the Ciliary Body of the Eye in Atg7 Tyr-Cre and Atg7 mice (Suppl. Fig. S5). Thus, Tyr- f/f Cre-mediated deletion of Atg7 caused aberrant accumulation of Atg7 Tyr-Cre Mice of the autophagy substrate p62 without deleterious effects on Outside of the brain, p62 accumulations were detected in skin the secretory function of neuroepithelial cells. melanocytes [18], choroid melanocytes of the eye, and retinal pigment epithelial cells [20]. However, the strongest accumu- Immunogold Electron Microscopy Shows Filamentous lation of p62, as judged from immunohistochemical and im- Structure of p62 Aggregates in Epithelial Cells munofluorescence analysis, was found in the epithelial cells of of the Ependyma and Neurons of Atg7f/f Tyr-Cre Mice the ciliary body of Atg7f/f Tyr-Cre mice (Fig. 6). p62 formed large inclusions in both pigmented and non-pigmented ciliary Next, we investigated the ultrastructural organization of the body epithelial cells of Atg7f/f Tyr-Cre mice. In approximately p62 accumulations in neurons and neuroepithelial cells of half of the Atg7f/f Tyr-Cre mice investigated, p62 accumula- Atg7f/f Tyr-Cre brain. Immunogold electron microscopy tions were also present at low abundance in the neuroretina Mol Neurobiol

Fig. 3 The accumulation of p62 is ubiquin p62 ubiquin p62 DNA only partially linked to accumulation of ubiquitin in A B C Atg7f/f Tyr-Cre mice. Brains of Atg7f/f (A–I) and Atg7f/f Tyr-Cre cortex (J–R) mice (age: 2 years) were sectioned and subjected to double immunolabeling for ubiquitin (red) and p62 (green). Nuclear DE F DNA was labeled with Hoechst f/f dye (blue). White arrows indicate ependyma examplary aggregates that were Atg7 positive for p62 and, weakly, ubiquitin; red arrows indicate an examplary aggregate that was GHI positive only for ubiquitin; and green arrows indicate an choroid plexus examplary aggregate that was positive only for p62. Scale bars, 50 μm JKL

cortex

MN O

Tyr-Cre ependyma f/f Atg7 PQR

choroid plexus

(Fig. 6). Eyes of Atg7f/f mice did not contain appreciable functional adapter protein [4, 33], our findings are likely rel- amounts of p62 (Fig. 6). evant for several aspects of neurobiology. In summary, both pigment cells expressing tyrosinase, such In the present study, the Tyr-Cre transgene [21, 23]was as melanocytes, and other neuroectodermal cells developed used to delete the floxed alleles of Atg7. The transgenic tyros- accumulations of p62 in response to Tyr-Cre-mediated dele- inase promoter was shown to be active in some neural crest tion of Atg7. This pattern suggested that the activation of the cell precursors of melanocytes, some smooth muscle cells of tyrosinase promoter in the Tyr-Cre transgene and the subse- the heart and cells of the enteric nervous system, but also in quent inactivation of Atg7 occurred in developmental precur- the brain [21, 23, 34, 35]. The neuronal expression of Tyr-Cre sor cells which inherit the lack of Atg7-dependent autophagy was used to delete phosphatase and tensin homolog (Pten)in to their neuronal and neuroepithelial progeny (Fig. 7). a subset of vagal neural crest cells, resulting in lethal intestinal pseudoobstruction of Tyr-Cre/° Ptenf/f mice [36]. We have previously investigated the effects of Tyr-Cre mediated dele- Discussion tion of Atg7 on skin melanocytes and retinal pigment epithe- lial cells [18, 20]. Atg7f/f Tyr-Cre mice displayed mild defects The results of this study establish Atg7f/f Tyr-Cre mice as a in hair pigmentation and alteration in the turnover of the model for the study of autophagy-deficiency in non-pigment C57BL/6 background-specific M450 variant of RPE65 and cells of the neuroectodermal lineage and support the hypoth- reached an age beyond 2 years. Of note, tyrosinase is not esis of filament formation of endogenous p62 as a mechanism involved in the synthesis of neuromelanin, the pigment within of sequestering p62 and possibly other proteins in autophagy- the substantia nigra, a region of the midbrain [37], and accord- deficient cells. Given the emerging role of p62 as a multi- ingly, a specific deletion of Atg7 was not expected in the Mol Neurobiol

Fig. 4 Immunogold electron microscopy shows filamentous structure of containing a p62 body (framed area). (B) Electron micrograph of the p62 aggregates in neurons and epithelial cells of the ependyma of Atg7f/f ependyma containing a p62 body (framed area). The framed regions of Tyr-Cre mice. Ultrathin sections of mouse brain were labeled with anti- panels A and B are shown at higher magnification in panels C and D. (E, p62 antibody conjugated to 10-nm gold particles. Electron micrographs F) The fine filamentous meshwork of a p62 body is shown at high mag- of only Atg7f/f Tyr-Cre brains (age: 23 months) are shown whereas Atg7f/f nifications. Lf, lipofuscin granules; M, mitochondrion; Ms, myelin brain (age: 26 months) showed only sparse immunogold labeling. (A) sheath; Mv, microvilli; Nu, nucleus; Syn, synapse. Scale bars: 1 μm Electron micrograph of the pericaryonal region of a thalamic neuron (A); 5 μm (B), 100 nm (C, E, F), 500 nm (D) substantia nigra of Atg7f/f Tyr-Cre mice. However, in line associated with abnormal hindlimb clasping reflexes at an with the reported expression of the Tyr-Cre transgene in age of 1.5 years and more in Atg7f/f Tyr-Cre mice (Suppl. multiple non-pigmented neuroectodermal cells, we found Fig. S1) but do not impair the survival of mice and therefore a diverse set of phenotypically abnormal cells. The pres- differ from the effects of Atg7 and Atg5 deletions in all neu- ent characterization of p62 aggregations in a series of rons which are lethal within 4–28 weeks after birth of mice [1, neuroectodermal lineage cells of Atg7f/f Tyr-Cre mice is 2]. Thus, Atg7f/f Tyr-Cre mice extend the list of viable mouse an important extension of previous studies because it sug- lines carrying deletions of Atg7 in distinct sets of brain neu- gests that cell-autonomous effects, such as those on me- rons [40–49]. Accordingly, Atg7f/f Tyr-Cre mice will be avail- lanocytes, may be accompanied by effects of altered sig- able for studying the impact of lack of autophagy and aberrant naling from neurons in Atg7f/f Tyr-Cre mice. accumulation of p62 in neurons of the brain in future studies. The accumulation of p62 in neurons and neuroepithelial Unexpectedly, neuroepithelial cells of the choroid plexus, cells of the brain and the eyes of Atg7f/f Tyr-Cre mice indicates the ependyma, and the ocular ciliary body were also affected that Atg7-dependent autophagy is abrogated in these cells. by Tyr-Cre-driven inactivation of Atg7. To the best of our The reduced degradation of p62 is an accepted in vivo marker knowledge, conditional suppression of autophagy has not of impaired autophagy [38, 39], especially when supported by been reported previously in these cell types. The structure of evidence for lack of transcriptional upregulation of Sqstm1/ p62 bodies in ependymal cells was similar to that in neurons, p62 expression (Suppl. Fig. S4). Importantly, the subcellular and the p62 aggregates in the ciliary body were even larger abnormalities in the brain of Atg7f/f Tyr-Cre mice are than those in neurons. The neuroepithelia of the choroid Mol Neurobiol

Fig. 5 Tyr-Cre-mediated deletion of Atg7 leads to alterations of the lipid composition in the brain. Lipids were extracted from the whole brain of Atg7f/f and Atg7f/f Tyr-Cre mice (age: 1 year) and analyzed as described in the Materials and Methods section. The abundance of individual lipids (A–F) was normalized to DPPC. The bars indicate quantities in arbitrary units and error bars indicate standard deviations. n = 4 per genotype. *, significant with P <0.05(two- sided t test). a.u., arbitrary units; Lyso-PPC, 1-palmitoyl-2- hydroxy-sn-glycero-3- phosphocholine; Lyso-SPC, 1- stearoyl-2-hydroxy-sn-glycero-3- phosphocholine; PAzPC, 1- palmitoyl-2-azelaoyl-sn-glycero- 3-phosphorylcholine; PLPC, 1- palmitoyl-2-linoleoyl-sn-glycero- 3-phosphorylcholine; SAzPC, 1- stearoyl-2-azelaoyl-sn-glycero-3- phosphorylcholine; SLPC, 1- stearoyl-2-linoleoyl-sn-glycero-3- phosphorylcholine

plexus and the ciliary body do not only share common em- Our data suggest that p62 forms filamentous aggregates bryological origins [50]; they also have similar functions. when sufficiently high-intracellular concentrations are reached Both of these epithelia secrete liquids, namely the cerebrospi- due to lack of autophagic turnover. In Atg7f/f Tyr-Cre mice, p62 nal fluid and the aqueous humor of the eye, respectively. As was diffusely distributed in melanocytes [18] and some brain the secretion rates of these epithelia control the intracranial neurons (Fig. 1) whereas in most of the affected neurons and pressure and the intraocular pressure, defects of the choroid neuroepithelial cells, p62 was concentrated in aggregates, also plexus epithelium and the ciliary body epithelium may be referred to as p62 bodies. The formation of p62 bodies has also medically relevant. It remains to be investigated whether been reported for other mouse models of autophagy deficiency Tyr-Cre-driven gene recombination can be used to study func- and therefore appears to be a characteristic feature of p62 [54]. tional parameters pertaining to the etiology of hydrocephalus Our investigation of the ultrastructural organization of p62 or glaucoma. Choroid plexus epithelial cells of Atg7f/f Tyr-Cre bodies demonstrates filaments of uniform thickness that were mice developed p62 bodies predominantly in the cell periph- densely bound by anti-p62 antibodies. Previous studies have ery, reminiscent of the so-called Biondi bodies or Biondi ring suggested that protein aggregates containing endogenous p62 tangles that appear during human aging and increase signifi- have a filamentous organization, e.g., in dendrites of dopami- cantly in patients with Alzheimer’sdisease[51, 52]. Although nergic neurons of mice that carry a tyrosine hydroxylase (TH) the structures of the p62 aggregates in the mutant mice and cell-specific deletion of Atg7 (Atg7f/f;TH-IRES-Cre)[43]. those of the tangles in humans are different [53], possible However, only recently evidence from studies involving re- similarities in mechanisms of formation should be combinant p62 in vitro [55, 56] and recombinant p62 investigated. expressed by an adeno-associated virus vector injected into Mol Neurobiol

Fig. 6 Tyr-Cre-mediated deletion Atg7f/f Atg7f/f Tyr-Cre of Atg7 leads to accumulation of p62 in the ciliary body of the eye. A B Eyes of Atg7f/f (A,C,E,G)and Atg7f/f Tyr-Cre (B,D,F,H)mice were sectioned and subjected to immunohistochemistry (IHC) for IHC: p62 ciliary body lens p62 (red) (A, B) and to double neurorena immunolabeling for ubiquitin ciliary body (red) and p62 (green) (C–H). neurorena Nuclear DNA was labeled with Hoechst dye (blue). Red arrow- heads (A, B) indicate p62-positive ciliary body epithelial cells and C D sparse p62-positive cells in the neuroretina of Atg7f/f Tyr-Cre mice. The images are representa- ubiquin tive for at least n = 3 per genotype and age group (10 months in panels A and B, and 21 months in C–H). Scale bars, 100 μm (A, B), 50 μm(C–H) E F

p62

G H

ubiquin p62 DNA

the rat substantia nigra [57] have suggested that p62 itself is Our demonstration of elevated amounts of the oxidized sufficient to form filaments. The structure and the dimensions phospholipids PAzPC and SAzPC in Atg7f/f Tyr-Cre brains of the p62 filaments in neurons and neuroepithelial cells of indicates that the turnover of these substances is altered. The Atg7f/f Tyr-Cre mice is similar or identical to those reported differential effects of autophagy inhibition on different classes for pure p62 [55] indicating that endogenous p62 is the main, of oxidized phospholipids (PAzPC and SAzPC versus Lyso- if not the only, component of these filaments. Our double- PPC, Lyso-SPC, PLPC, and SLPC in the Atg7f/f Tyr-Cre immunofluorescence labeling results suggest that ubiquitin is brain), which are similar to those observed in Atg7f/f K14- also present at elevated concentrations in p62 bodies of Atg7f/f Cre skin cells [58], may be caused by the delivery of sub- Tyr-Cre mice; however, the relative abundance inside versus strates to lysosomal phospholipase A2 [59]. Augmentation outside of these aggregates was much lower for ubiquitin than of oxidatively modified phospholipids in cellular membranes p62. Furthermore, besides ubiquitin-positive aggregates, ap- affects their polarity and permeability, which has been specif- parently ubiquitin-negative aggregates were detected, suggest- ically demonstrated for PLPC-derived PazPC [60]. ing that ubiquitinated proteins do not play an essential role in Additionally, PAzPC has been identified as a chemical modi- p62 body formation. Thus, we propose that the formation of fier of proteins in oxidative stress [32], and as inducer of p62 bodies in Atg7f/f Tyr-Cre mice is likely driven by an inher- amyloid fibril aggregation [61]. Further research will address ent tendency to filamentous polymerization of p62 at elevated possible mechanistic links between phospholipid oxidation concentrations. and the protein aggregates in the brain. Mol Neurobiol

Atg7f/f Tyr-Cre Open Access This article is distributed under the terms of the Creative vs. Atg7f/f Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, Skin melanocytes p62 ↑ distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link Pigment cells to the Creative Commons license, and indicate if changes were made. Renal pigment p62 ↑ epithelial cells References Choroid plexus p62 –/↑ epithelium 1. Hara T, Nakamura K, Matsui M, Yamamoto A, Nakahara Y, Neuroepithelial Suzuki-Migishima R, Yokoyama M, Mishima K et al (2006) Ependyma p62 –/↑ cells Suppression of basal autophagy in neural cells causes neurodegen- erative disease in mice. Nature 441:885–889 Ciliary body 2. Komatsu M, Waguri S, Chiba T, Murata S, Iwata J, Tanida I, Ueno p62 ↑ epithelium T, Koike M et al (2006) Loss of autophagy in the central nervous system causes neurodegeneration in mice. Nature 441:880–884 3. Mariño G, Madeo F, Kroemer G (2011) Autophagy for tissue ho- Brain neurons p62 –/↑ meostasis and neuroprotection. Curr Opin Cell Biol 23:198–206 4. Katsuragi Y, Ichimura Y, Komatsu M (2015) p62/SQSTM1 func- tions as a signaling hub and an autophagy adaptor. FEBS J 282: Other 4672–4678 neuroectoderm- p62 ? 5. Mizushima N, Komatsu M (2011) Autophagy: renovation of cells derived cells and tissues. Cell 147:728–741 6. Viscomi MT, D'Amelio M (2012) The BJanus-faced role^ of au- Fig. 7 Summary of differential effects of Tyr-Cre-mediated deletion of tophagy in neuronal sickness: focus on neurodegeneration. Mol Atg7 on p62 abundance in neuroectoderm-derived cell types. Neurobiol 46:513–521 f/f f/f Neuroectoderm-derived cells of Atg7 Tyr-Cre and Atg7 (fully 7. Hu Z, Yang B, Mo X, Xiao H (2015) Mechanism and regulation of autophagy-competent) mice were characterized with regard to the abun- autophagy and its role in neuronal diseases. Mol Neurobiol 52: dance and distribution of p62. The results of previous studies on melano- 1190–1209 cytes and retinal pigment epithelial cells [18, 20] and of the present study 8. Tschachler E, Eckhart L (2017) Autophagy - how to control your are schematically summarized. ↑, accumulation; −,absent;−/↑, absent intracellular diet. Br J Dermatol 176:1417–1419 and accumulated in different subsets of this cell type 9. Zatloukal K, Stumptner C, Fuchsbichler A, Heid H, Schnoelzer M, Kenner L, Kleinert R, Prinz M et al (2002) p62 is a common com- In summary, our results suggest that Tyr-Cre-mediated sup- ponent of cytoplasmic inclusions in protein aggregation diseases. Am J Pathol 160:255–263 pression of Atg7-dependent autophagy causes the aggregation 10. Wong E, Cuervo AM (2010) Autophagy gone awry in neurodegen- of p62 in neurons and neuroepithelial cells and an increase of erative diseases. Nat Neurosci 13:805–811 oxidized lipid species in the brain. Because of the apparently 11. Yamamoto A, Yue Z (2014) Autophagy and its normal and patho- slow accumulation of damage, Atg7f/f Tyr-Cre mice can be used genic states in the brain. Annu Rev Neurosci 37:55–78 to study the effects of p62 aggregation in neuroectoderm- 12. Nikoletopoulou V, Papandreou ME, Tavernarakis N (2015) Autophagy in the physiology and pathology of the central nervous derived cells during aging. system. Cell Death Differ 22:398–407 13. Bové J, Martínez-Vicente M, Vila M (2011) Fighting neurodegen- Acknowledgements Open access funding provided by Medical eration with rapamycin: mechanistic insights. Nat Rev Neurosci 12: University of Vienna. This project was supported by the 437–452 Herzfelder’sche Familienstiftung. We thank Masaaki Komatsu (Tokyo 14. Srivastava A, Kumar V, Pandey A, Jahan S, Kumar D, Rajpurohit Metropolitan Institute of Medical Science, Tokyo, Japan) for providing CS, Singh S, Khanna VK et al (2017) Adoptive autophagy activa- Atg7-floxed mice, Johannes Hainfellner and Andreas Pollreisz for helpful tion: a much-needed remedy against chemical induced discussions, and Ferencne Truszka for the technical assistance in electron neurotoxicity/developmental neurotoxicity. Mol Neurobiol 54: microscopy. The financial support of the Federal Ministry for Digital and 1797–1807 Economic Affairs (BMDW) of Austria and the National Foundation for 15. Sukseree S, Rossiter H, Mildner M, Pammer J, Buchberger M, Research, Technology, and Development of Austria to the Christian Gruber F, Watanapokasin R, Tschachler E et al (2013) Targeted de- Doppler Laboratory for Biotechnology of Skin Aging is gratefully ac- letion of Atg5 reveals differential roles of autophagy in keratin K5- knowledged. LE and FG are participants of the COST Action CA15138, expressing epithelia. Biochem Biophys Res Commun 430:689–694 European Network of Multidisciplinary Research and Translation of 16. Kuma A, Hatano M, Matsui M, Yamamoto A, Nakaya H, Autophagy knowledge (TRANSAUTOPHAGY) that is supported by Yoshimori T, Ohsumi Y, Tokuhisa T et al (2004) The role of au- the European Union Framework Programme Horizon 2020. tophagy during the early neonatal starvation period. Nature 432: 1032–1036 Compliance with Ethical Standards 17. Komatsu M, Waguri S, Ueno T, Iwata J, Murata S, Tanida I, Ezaki J, Mizushima N et al (2005) Impairment of starvation-induced and constitutive autophagy in Atg7-deficient mice. J Cell Biol 169: Conflict of Interest The authors declare that they have no conflict of 425–434 interest. Mol Neurobiol

18. Zhang CF, Gruber F, Ni C, Mildner M, Koenig U, Karner S, Barresi A subpopulation of smooth muscle cells, derived from melanocyte- C, Rossiter H et al (2015) Suppression of autophagy dysregulates competent precursors, prevents patent ductus arteriosus. PLoS One the antioxidant response and causes premature senescence of mela- 8:e53183. nocytes. J Invest Dermatol 2015 135:1348–1357 36. Puig I, Champeval D, De Santa BP, Jaubert F, Lyonnet S, Larue L 19. Ni C, Narzt MS, Nagelreiter IM, Zhang CF, Larue L, Rossiter H, (2009) Deletion of Pten in the mouse enteric nervous system in- Grillari J, Tschachler E et al (2016) Autophagy deficient melanocytes duces ganglioneuromatosis and mimics intestinal display a senescence associated secretory phenotype that includes pseudoobstruction. J Clin Invest 119:3586–3596 oxidized lipid mediators. Int J Biochem Cell Biol 2016 81:375–382 37. Zecca L, Tampellini D, Gerlach M, Riederer P, Fariello RG, Sulzer 20. Sukseree S, Chen YT, Laggner M, Gruber F, Petit V, Nagelreiter D (2001) Substantia nigra neuromelanin: structure, synthesis, and IM, Mlitz V, Rossiter H et al (2016) Tyrosinase-Cre-mediated de- molecular behaviour. Mol Pathol 54:414–418 letion of the autophagy gene Atg7 leads to accumulation of the 38. Klionsky et al. (2016) Guidelines for the use and interpretation of RPE65 variant M450 in the retinal pigment epithelium of C57BL/ assays for monitoring autophagy (3rd edition). Autophagy 12:1–222. 6 mice. PLoS One 2016 11:e0161640 39. Galluzzi L, Bravo-San Pedro JM, Levine B, Green DR, Kroemer G 21. Delmas V, Martinozzi S, Bourgeois Y, Holzenberger M, Larue L (2017) Pharmacological modulation of autophagy: therapeutic po- (2003) Cre-mediated recombination in the skin melanocyte lineage. tential and persisting obstacles. Nat Rev Drug Discov 16:487–511 – Genesis 36:73 80 40. Kim HJ, Cho MH, Shim WH, Kim JK, Jeon EY,Kim DH, Yoon SY 22. Tonks ID, Nurcombe V, Paterson C, Zournazi A, Prather C, Mould (2016) Deficient autophagy in microglia impairs synaptic pruning AW, Kay GF (2003) Tyrosinase-Cre mice for tissue-specific gene and causes social behavioral defects. Mol Psychiatry. https://doi. ablation in neural crest and neuroepithelial-derived tissues. Genesis org/10.1038/mp.2016.103 – 37:131 138 41. Komatsu M, Wang QJ, Holstein GR, Friedrich VL Jr, Iwata J, 23. Colombo S, Petit V, Kumasaka M, Delmas V, Larue L (2007) Kominami E, Chait BT, Tanaka K et al (2007) Essential role for Flanking genomic region of Tyr:Cre mice, rapid genotyping for autophagy protein Atg7 in the maintenance of axonal homeostasis – homozygous mice. Pigment Cell Res 20:305 306 and the prevention of axonal degeneration. Proc Natl Acad Sci U S 24. Fischer H, Rossiter H, Ghannadan M, Jaeger K, Barresi C, Declercq A 104:14489–14494 W, Tschachler E, Eckhart L (2005) Caspase-14 but not caspase-3 is 42. Coupé B, Ishii Y, Dietrich MO, Komatsu M, Horvath TL, Bouret processed during the development of fetal mouse epidermis. – SG (2012) Loss of autophagy in pro-opiomelanocortin neurons Differentiation 73:406 413 perturbs axon growth and causes metabolic dysregulation. Cell 25. Gruber F, Bicker W, Oskolkova OV, Tschachler E, Bochkov VN Metab 15:247–255 (2012) A simplified procedure for semi-targeted lipidomic analysis 43. Friedman LG, Lachenmayer ML, Wang J, He L, Poulose SM, of oxidized phosphatidylcholines induced by UVA irradiation. J Komatsu M, Holstein GR, YueZ (2012) Disrupted autophagy leads Lipid Res 53:1232–1242 to dopaminergic axon and dendrite degeneration and promotes pre- 26. DeMattos RB, Bales KR, Parsadanian M, O'Dell MA, Foss EM, synaptic accumulation of α-synuclein and LRRK2 in the brain. J Paul SM, Holtzman DM (2002) Plaque-associated disruption of Neurosci 32:7585–7593 CSF and plasma amyloid-beta (Abeta) equilibrium in a mouse mod- 44. Torres CA, Sulzer D (2012) Macroautophagy can press a brake on el of Alzheimer’s disease. J Neurochem 81:229–236 presynaptic neurotransmission. Autophagy 8:1540–1541 27. Eckhart L, Ballaun C, Uthman A, Kittel C, Stichenwirth M, Buchberger M, Fischer H, Sipos W et al (2005) Identification and 45. Inoue K, Rispoli J, Yang L, Macleod D, Beal MF, Klann E, characterization of a novel mammalian caspase with proapoptotic Abeliovich A (2013) Coordinate regulation of mature dopaminer- activity. J Biol Chem 280:35077–35080 gic axon morphology by macroautophagy and the PTEN signaling 28. Mangiarini L, Sathasivam K, Seller M, Cozens B, Harper A, pathway. PLoS Genet 9:e1003845 Hetherington C, Lawton M, Trottier Y et al (1996) Exon 1 of the 46. Cho MH, Cho K, Kang HJ, Jeon EY,Kim HS, Kwon HJ, Kim HM, HD gene with an expanded CAG repeat is sufficient to cause a Kim DH et al (2014) Autophagy in microglia degrades extracellular β progressive neurological phenotype in transgenic mice. Cell 87: -amyloid fibrils and regulates the NLRP3 inflammasome. – 493–506 Autophagy 10:1761 1775 29. Sakka L, Coll G, Chazal J (2011) Anatomy and physiology of 47. Nilsson P, Sekiguchi M, Akagi T, Izumi S, Komori T, Hui K, cerebrospinal fluid. Eur Ann Otorhinolaryngol Head Neck Dis Sörgjerd K, Tanaka M et al (2015) Autophagy-related protein 7 β β 128:309–316 deficiency in amyloid (A ) precursor protein transgenic mice β β 30. Singh R, Kaushik S, Wang Y, Xiang Y, Novak I, Komatsu M, decreases A in the multivesicular bodies and induces A accu- – Tanaka K, Cuervo AM et al (2009) Autophagy regulates lipid me- mulation in the Golgi. Am J Pathol 185:305 313 tabolism. Nature:4581131–4581135 48. Niu XY, Huang HJ, Zhang JB, Zhang C, Chen WG, Sun CY, Ding 31. Kaur J, Debnath J (2015) Autophagy at the crossroads of catabo- YQ, Liao M (2016) Deletion of autophagy-related gene 7 in dopa- lism and anabolism. Nat Rev Mol Cell Biol 16:461–472 minergic neurons prevents their loss induced by MPTP. 32. Januszewski AS, Alderson NL, Jenkins AJ, Thorpe SR, Baynes JW Neuroscience 339:22–31 (2005) Chemical modification of proteins during peroxidation of 49. Sato S, Uchihara T, Fukuda T, Noda S, Kondo H, Saiki S, Komatsu phospholipids. J Lipid Res 46:1440–1449 M, Uchiyama Y et al (2018) Loss of autophagy in dopaminergic 33. Lee H, Ahn HH, Lee W, Oh Y, Choi H, Shim SM, Shin J, Jung YK neurons causes Lewy pathology and motor dysfunction in aged (2016) ENC1 modulates the aggregation and neurotoxicity of mu- mice. Sci Rep 8:2813 tant huntingtin through p62 under ER stress. Mol Neurobiol 53: 50. Janssen SF, Gorgels TG, Ten Brink JB, Jansonius NM, Bergen AA 6620–6634 (2014) -based comparison of the human secretory 34. Puig I, Yajima I, Bonaventure J, Delmas V, Larue L (2009) The neuroepithelia of the brain choroid plexus and the ocular ciliary body: tyrosinase promoter is active in a subset of vagal neural crest cells potential implications for glaucoma. Fluids Barriers CNS 11:2 during early development in mice. Pigment Cell Melanoma Res 22: 51. Wen GY, Wisniewski HM, Kascsak RJ (1999) Biondi ring tangles 331–334 in the choroid plexus of Alzheimer's disease and normal aging 35. Yajima I, Colombo S, Puig I, Champeval D, Kumasaka M, Belloir brains: a quantitative study. Brain Res 832:40–46 E, Bonaventure J, Mark M, Yamamoto H, Taketo MM, Choquet P, 52. Wolburg H, Paulus W (2010) Choroid plexus: biology and pathol- Etchevers HC, Beermann F, Delmas V, Monassier L, Larue (2013) ogy. Acta Neuropathol 119:75–88 Mol Neurobiol

53. Eriksson L, Westermark P (1990) Characterization of intracellular 58. Zhao Y, Zhang CF, Rossiter H, Eckhart L, König U, Karner S, amyloid fibrils in the human choroid plexus epithelial cells. Acta Mildner M, Bochkov VN et al (2013) Autophagy is induced by Neuropathol 80:597–603 UVA and promotes removal of oxidized phospholipids and protein 54. Komatsu M, Waguri S, Koike M, Sou YS, Ueno T, Hara T, aggregates in epidermal keratinocytes. J Invest Dermatol 133: Mizushima N, Iwata J et al (2007) Homeostatic levels of p62 con- 1629–1637 trol cytoplasmic inclusion body formation in autophagy-deficient 59. Abe A, Hiraoka M, Ohguro H, Tesmer JJ, Shayman JA (2017) mice. Cell 131:1149–1163 Preferential hydrolysis of truncated oxidized glycerophospholipids 55. Ciuffa R, Lamark T, Tarafder AK, Guesdon A, Rybina S, Hagen WJ, by lysosomal phospholipase A2. J Lipid Res 58:339–349 Johansen T, Sachse C (2015) The selective autophagy receptor p62 60. Volinsky R, Kinnunen PK (2013) Oxidized phosphatidylcholines in forms a flexible filamentous helical scaffold. Cell Rep 11:748–758 membrane-level cellular signaling: from biophysics to physiology 56. Johansen T, Sachse C (2015) The higher-order molecular organiza- and molecular pathology. FEBS J 280:2806–2816 – tion of p62/SQSTM1. Oncotarget 6:16796 16797 61. Domanov YA, Kinnunen PK (2008) Islet amyloid polypeptide 57. Jackson KL, Lin WL, Miriyala S, Dayton RD, Panchatcharam M, forms rigid lipid-protein amyloid fibrils on supported phospholipid McCarthy KJ, Castanedes-Casey M, Dickson DW et al (2017) p62 bilayers. J Mol Biol 376:42–54 pathology model in the rat substantia nigra with filamentous inclu- sions and progressive neurodegeneration. PLoS One 12:e0169291